The End of College

1. The American College: Grotesquely Expensive, Shamefully Indifferent.

“American higher education,” they wrote, “is characterized by limited or no learning for a large proportion of students.”

Skyrocketing costs. College tuition has tripled since the 1980s, far outpacing family income, forcing students into massive debt, now exceeding $1 trillion. This financial burden prevents homeownership and leads to high default rates, trapping many without the promised economic mobility.

Poor outcomes. Despite the cost, graduation rates are low, with less than two-thirds finishing a four-year degree within six years, and community college rates are even worse. Studies show many graduates lack basic literacy and critical thinking skills, indicating limited learning gains during college. This suggests the value proposition is failing for many.

Declining rigor. Colleges demand less study time from students than in the past, while simultaneously inflating grades. This lack of challenge, coupled with high dropout rates and low learning gains, reveals a system that is failing to educate effectively, making the exorbitant cost even harder to justify.

2. A Historical Accident: The Flawed Hybrid University.

The modern American university has endured for this long only because, almost as soon as it was created, it began an epic run of good luck that is only now coming to an end.

Three competing ideas. The modern American university emerged in the late 19th century by combining three conflicting purposes: practical training (land-grant), research (German model), and liberal arts (Newman). Charles Eliot of Harvard reconciled these by creating the elective system and requiring bachelor's degrees for graduate school, but this led to structural incoherence.

Research prioritized. The German research model won the internal battle, making the PhD the standard for faculty, despite it not indicating teaching ability. Professors were rewarded for research, not teaching, leading to widespread indifference towards undergraduate instruction, a flaw pointed out by critics like William James and Jacques Barzun.

Survival by luck. This flawed model survived due to external factors: America's rise as a global power, the GI Bill, massive federal research funding during the Cold War, increasing demand for college degrees in the information economy, and a regulatory system controlled by existing institutions that stifled competition.

3. Technology's Promise: From Books to AI Tutors.

Soon... millions of schoolchildren will have access to what Philip of Macedon’s son Alexander enjoyed as a royal prerogative: the personal services of a tutor as well-informed and responsive as Aristotle.

Learning is active. Education is not passive reception but active construction of knowledge by integrating new information into existing neural patterns. This requires hard work, like solving problem sets, which strengthens understanding and builds expertise.

Personalized learning needed. Because each person's neural patterns are unique, effective education requires individualization, historically achieved through tutoring. Early attempts to use technology like mail, radio, and TV for "distance education" failed to fundamentally change teaching because they only transmitted information, they didn't personalize learning.

Computers process, not just transmit. Patrick Suppes and others saw that computers could analyze student performance and adapt instruction, simulating a tutor. This ability to process information and respond to individual needs, based on theories of cognition and AI, is the key difference technology brings to education.

4. The Internet Changes Everything (Eventually).

When the Internet became available to ordinary people in the mid-1990s, many of these barriers began to fall.

Early attempts failed. Despite predictions, early internet education ventures like Columbia's Fathom.com failed because technology was immature and they misunderstood what people were buying: not just content, but credentials and the "experience" of being at a prestigious university. They tried to replicate the old model online.

Technology improved. Over time, mobile computing, cloud storage, and social networking matured, making online interaction and access to powerful computing much cheaper and more ubiquitous. This laid the groundwork for more sophisticated online learning environments.

Colleges resisted change. Traditional universities largely ignored these technological shifts in teaching, adding "tech fees" while raising tuition. Protected by accreditation monopolies and cultural inertia, they avoided the productivity gains technology offered, unlike other information-based industries.

5. Thunder Lizards and Platforms: Silicon Valley Attacks.

Companies built for such disruption... are like thunder lizards.

Software eats the world. Silicon Valley entrepreneurs see education as a massive, analog market ripe for disruption by software and technology platforms. They are fueled by venture capital, ambition, and a belief that no incumbent can withstand exponential technological progress forever.

Unbundling the university. Start-ups are targeting specific functions of the hybrid university, offering services like:

• Textbook rentals (Chegg)
• Online study groups (Piazza)
• Coaching (InsideTrack)
• Job-specific training (Dev Bootcamp)
• Alternative credentials (Open Badges, Accredible)

Platforms scale rapidly. Companies like Coursera aim to build dominant online education platforms, leveraging network effects. They don't create content themselves but provide the infrastructure for others (like universities) to offer courses, aiming for massive scale and low marginal costs.

6. Elite Universities Join the Fray: MOOCs Emerge.

Harvard and MIT held a rare joint press conference to announce the creation of a new venture called edX.

Stanford's accidental MOOC. Sebastian Thrun and Peter Norvig put their Stanford AI course online for free, attracting over 100,000 students. This "massive open online course" (MOOC) phenomenon, though not technologically revolutionary, was significant because it came from a top-tier university.

Disrupting the brand. Thrun's MOOC challenged the idea that elite education was tied to a physical place and expensive tuition. When Stanford resisted offering credit, Thrun left to start Udacity, a for-profit MOOC company.

Fear of being left behind. The success of Stanford's MOOC and the launch of Udacity and Coursera panicked other elite universities. Driven by status competition, institutions like Harvard and MIT quickly launched edX to establish their presence in the online space, lending their valuable brands to the MOOC movement.

7. Rational Education: Data, AI, and Personalized Learning.

Rational education is emerging from profound changes of scale...

Learning generates data. Digital learning environments create vast amounts of data on student behavior, progress, and difficulties. This data allows for unprecedented analysis of how people learn, moving education from anecdote to data science.

AI personalizes instruction. Cognitive tutors and AI models, like those developed at Carnegie Mellon, can diagnose individual student needs and adapt the learning experience in real-time. This provides personalized support, much like a human tutor, but at scale.

Machine learning optimizes. Analyzing large datasets using machine learning can reveal patterns in learning that human researchers might miss. This allows for continuous improvement and optimization of educational designs, making them more effective over time.

8. Beyond the Diploma: The Rise of Open Credentials.

Mozilla’s Project, called Open Badges, is designed to create an alternative system of credentials that is controlled by no central authority.

Diplomas are outdated. Traditional college degrees and transcripts are poor signals of actual knowledge and skills. They are time-based, opaque, easily forged, and controlled by institutions, not learners. They primarily signal entry into an elite group or completion of a bureaucratic process.

Digital credentials offer more. New systems like Open Badges and Accredible allow individuals to collect, control, and display verifiable evidence of their learning from various sources. These digital credentials can include detailed metadata, portfolios of work, and endorsements.

Machine discoverability. These new credentials are designed to be easily found and understood by machines, allowing employers and other organizations to search for specific skills and knowledge. This bypasses the traditional degree as a gatekeeper and opens opportunities for those without conventional credentials.

9. The Weight of Large Numbers: Billions Need Education.

The number of additional people who will want a college education over the next twenty years could exceed the number of people who have ever been to college in all of human history.

Global middle class growth. Billions of people are rising out of poverty, particularly in developing nations. This emerging global middle class sees education as the key to a better life for their children and will demand higher learning opportunities.

Hybrid model cannot scale. The traditional, expensive, place-bound hybrid university model cannot possibly accommodate this massive increase in demand. Building enough physical campuses and training enough PhD faculty is financially and logistically impossible on this scale.

Technology enables access. Ubiquitous, cheap computing and telecommunications networks mean billions will have access to online learning. MOOCs and other digital platforms are already seeing massive enrollment from around the world, demonstrating the huge pent-up demand for accessible higher education.

10. The University of Everywhere: A New Landscape.

The future of higher education is one in which educational organizations shrink back to a human scale.

Free courses, cheap assessment. The cost of providing digital course content is near zero. While assessment and human interaction will still cost money, the overall price of education will fall dramatically, making high-quality learning affordable for billions.

New learning organizations. The low cost of creating digital learning environments will lead to a proliferation of new educational organizations. These may be physical spaces, like 21st-century Carnegie Libraries, focused on mentoring and community, or purely virtual entities.

Specialization and collaboration. These new organizations will specialize in different aspects of learning (e.g., specific subjects, types of support, pedagogical philosophies). They will collaborate using open-source tools and platforms, creating a diverse and interconnected ecosystem of learning opportunities.

11. Navigating the Future: Advice for Students and Parents.

What parents can do is to help their children build the intellectual and emotional tools they will need for these demanding and rewarding tasks.

Choose wisely now. If attending a traditional college, prioritize institutions with clear educational philosophies and rigorous programs, not just prestige or amenities. Avoid paying high tuition for services available cheaply or free online.

Focus on learning, not just credentials. Regardless of the institution, prioritize deep learning and skill acquisition, especially in challenging arts, sciences, math, and engineering fields. The value of generic degrees will decline as open credentials rise.

Prepare for lifelong learning. The future requires continuous learning. Embrace digital learning environments and new credentialing systems. Build a personal educational identity by collecting verifiable evidence of skills and knowledge gained throughout life.

Cultivate grit and curiosity. The University of Everywhere will be demanding. Success requires self-motivation, discipline, and a genuine love of learning. Help children develop these traits, as they will be more important than ever in a world of abundant, challenging educational opportunities.

Last updated: June 8, 2025